SUBPROJECT 5: BEHAVIOR CHANGE IN THE SAVINGS AND HEALTH DOMAINS This subproject involves a series of behavioral experiments designed to improve well-being. Past research has shown that framing and other kinds of "choice architecture" can have a large impact on economic outcomes. This is true even when the stakes are high. However, researchers and policy-makers still have only a rudimentary understanding of how and why these framing manipulations work. The behavioral experiments we propose will advance our understanding of these mechanisms. The experiments will evaluate decision-making in the domains of retirement savings and health, both areas with important consequences for social welfare. In the domain of retirement savings, we will evaluate how choice architecture impacts savings plan participation, savings rates, and annuitization choices at the time of retirement. In the health domain, we will evaluate how choice architecture impacts the utilization of preventative and diagnostic health services. The proposed research will (i) develop interventions for behavior change that build on successful previous interventions and that are inexpensive and scalable, and (ii) study and refine the mechanisms that made those previous interventions work. The planned research will help organizations and policymakers better understand how to use choice architecture when designing decision-making environments. In addition, the proposed research will help identify the general economic and psychological principles that underlie the operation of these framing effects.